NAME

Krang::Markup - Base class for browser-specific WYSIWYG HTML filtering

SYNOPSIS

Krang::Markup - Base class for browser-specific WYSIWYG HTML filtering

DESCRIPTION

Different browsers use different HTML tags to manage basic markup like bold, italic, underline, strike-through, subscript and superscript. To normalize the database content across the usage of different browsers, the HTML must be filtered accordingly when going to and coming from browser WYSIWYG areas.

Definition

A ``normalized tag'' is the HTML tag stored in the database and published on the net.

Example

Let's take BOLD text as an example. The normalized tag is STRONG.

Boldifying text in IE effectively inserts the STRONG tag.

Gecko, however, inserts the B tag, and WebKit wraps the text with a SPAN tag having its style attribute set to font-weight: bold.

When going to Gecko or WebKit the STRONG tag therefore has to be replaced with what the WYSIWYG commands of those browsers understand. And when coming from them, the normalized version has to be restored.

This module provides mockups for methods accomplishing this task.

INTERFACE

Subclasses must implement the following class methods:

• pkg('Markup::Subclass')->db2browser_map()
  
This method must return a list mapping normalized tags to their browser-specific equivalent. It is to be called by db2browser().
• pkg('Markup::Subclass')->browser2db_map()
  
This method must return a list mapping browser-specific HTML tags to their normalized equivalent. It is to be called by browser2db().
• pkg('Markup::Subclass')->db2browser(html => $html)
  
This method replaces normalized tags with their browser-specific equivalent using the mappings provided by db2browser_map().

It is passed the normalized HTML and returns a string with those mappings applied to it.

Due to strangeness of the internally used HTML::Element module this method must not directly return the modified HTML. Instead return the HTML returned by $pkg->tidy_up_after_treebuilder(tree => $tree) - the argument $tree being a HTML::TreeBuilder object.

• pkg('Markup::Subclass')->browser2db(html => $html)
  
This method replaces browser-specific tags with their normalized equivalent according to the mappings provided by browser2db_map()

It is passed the HTML coming from the browser and returns a string with those mappings applied.

Due to strangeness of the internally used HTML::Element module this method must not directly return the modified HTML. Instead return the HTML returned by $pkg->tidy_up_after_treebuilder(tree => $tree) - the argument $tree being a HTML::TreeBuilder object.

• pkg('Markup::Subclass')->tidy_up_after_treebuilder($html_tree_object)
  
This method should be called at the end of the two mapping methods db2browser() and browser2db(). It chops the BODY tag HTML::TreeBuilder wraps around the passed-in HTML, removes the trailing newline added by HTML::Element's as_HTML(), destroys the tree object and finally returns the HTML.
• pkg('Markup::Subclass')->remove_junk(\$html)
  
This method does:

 remove tags w/o content inside
 remove adjacent closing/opening tags while preserving whitespace in between
 remove excess whitespace
 remove leading whitespace
 remove trailing whitespace
 remove leading BR tags
 remove trailing BR tags

It must be passed a scalar reference to a string containing the HTML to be cleaned.